<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="luo-etal-2026-teaching">
    <titleInfo>
        <title>Teaching by Learning: Harnessing LLMs Collaboration for Adaptive Data Synthesis via Cognitive Load Feedback</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yue</namePart>
        <namePart type="family">Luo</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Feng</namePart>
        <namePart type="family">Zhao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yu</namePart>
        <namePart type="family">Yang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Guandong</namePart>
        <namePart type="family">Xu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Recent advancements in Large Language Models (LLMs) are increasingly driven by using LLMs themselves to synthesize large amounts of instruction-tuning data. To transcend the limitations of single model synthesis, recent research has shifted toward multiple model data synthesis, where distinct models assume specialized roles to synthesize data through interaction. However, existing feedback cannot distinguish truly high-quality instructions, those that promote complex reasoning, deviate from the learned distribution, and require multi-source knowledge, due to coarse granularity and signal dilution. In human cognition, truly high-quality instructions require greater mental effort—the **Cognitive Load**—to address. In this paper, we propose a paradigm shift in feedback. Instead of external, result-oriented evaluation, we propose an internal, process-aware feedback based on target models’ computational effort. The novel feedback, namely **Inference Stress Index (ISI)**, quantifies instruction quality in a fine-grained, dilution-resistant manner by focusing on high-uncertainty tokens. Building on this, we present the **Co**gnitive-**F**eedback **A**daptive **G**eneration (**CoFAG**) framework, a self-evolving multiple model collaboration paradigm involving a synthesizer model, a fixer model, and a collaborator model. Extensive experiments demonstrate that CoFAG significantly outperforms other data synthesis techniques on complex reasoning benchmarks, validating the superiority of using internal cognitive feedback for generating high-quality data.</abstract>
    <identifier type="citekey">luo-etal-2026-teaching</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.1214/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>27739</start>
            <end>27754</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Teaching by Learning: Harnessing LLMs Collaboration for Adaptive Data Synthesis via Cognitive Load Feedback
%A Luo, Yue
%A Zhao, Feng
%A Yang, Yu
%A Xu, Guandong
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F luo-etal-2026-teaching
%X Recent advancements in Large Language Models (LLMs) are increasingly driven by using LLMs themselves to synthesize large amounts of instruction-tuning data. To transcend the limitations of single model synthesis, recent research has shifted toward multiple model data synthesis, where distinct models assume specialized roles to synthesize data through interaction. However, existing feedback cannot distinguish truly high-quality instructions, those that promote complex reasoning, deviate from the learned distribution, and require multi-source knowledge, due to coarse granularity and signal dilution. In human cognition, truly high-quality instructions require greater mental effort—the **Cognitive Load**—to address. In this paper, we propose a paradigm shift in feedback. Instead of external, result-oriented evaluation, we propose an internal, process-aware feedback based on target models’ computational effort. The novel feedback, namely **Inference Stress Index (ISI)**, quantifies instruction quality in a fine-grained, dilution-resistant manner by focusing on high-uncertainty tokens. Building on this, we present the **Co**gnitive-**F**eedback **A**daptive **G**eneration (**CoFAG**) framework, a self-evolving multiple model collaboration paradigm involving a synthesizer model, a fixer model, and a collaborator model. Extensive experiments demonstrate that CoFAG significantly outperforms other data synthesis techniques on complex reasoning benchmarks, validating the superiority of using internal cognitive feedback for generating high-quality data.
%U https://aclanthology.org/2026.emnlp-main.1214/
%P 27739-27754

Markdown (Informal)

[Teaching by Learning: Harnessing LLMs Collaboration for Adaptive Data Synthesis via Cognitive Load Feedback](https://aclanthology.org/2026.emnlp-main.1214/) (Luo et al., EMNLP 2026)

• Teaching by Learning: Harnessing LLMs Collaboration for Adaptive Data Synthesis via Cognitive Load Feedback (Luo et al., EMNLP 2026)

ACL

• Yue Luo, Feng Zhao, Yu Yang, and Guandong Xu. 2026. Teaching by Learning: Harnessing LLMs Collaboration for Adaptive Data Synthesis via Cognitive Load Feedback. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 27739–27754, Budapest, Hungary. Association for Computational Linguistics.